Growing environmental data needs in the Anthropocene: scaling up data flow from amateurs and experts through cyberinfrastructure

The recent period of human-driven environmental change has been so rapid and different than any other in Earth’s history that we now recognize it as a new era - the Anthropocene. Managing these changes to conserve... [ view full abstract ]

The recent period of human-driven environmental change has been so rapid and different than any other in Earth’s history that we now recognize it as a new era - the Anthropocene. Managing these changes to conserve biodiversity is a priority for our planet’s health, yet it requires a broad geographic and temporal context that scales from local to regional to continental to global. For many researchers, citizen science is seen as the best solution for collecting data at the temporal and spatial scale needed to address these global problems. However, managing such large data flows from volunteers challenges the technologies typically developed for local studies, especially when extra steps are needed to build in quality control of information collected by non-experts. We present the cyberinfrastructure of eMammal, a large-scale mammal population study, as a case study for data management and workflow that meets the needs of both small and large scale research questions. eMammal recruited more than 500 volunteers to deploy remote cameras in forests across six states in the Mid-Atlantic region. Volunteers collected over 2.6 million images in a two-year period by setting motion-sensitive camera traps at more than 2,300 locations. We have implemented a cloud computing workflow to manage data, including a central website to coordinate volunteers, custom software for remote photo tagging and uploading, and storage and curation of photos and meta-data in a Smithsonian digital repository.The workflow includes an ‘expert review’ process to verify all data from volunteers, and we continue to develop automated image analysis to make image processing more efficient. We conclude by proposing cyberinfrastructure solutions that would allow the project to reach a global scale with 30,000 or even 100,000 camera trap deployments.